What is ortho generation? A beginner’s guide to making ortho images from drone photos

Table of contents

• What is an ortho image?

• Benefits of creating ortho images with drones

• Equipment and software required for ortho generation

• Steps to create ortho images

• Tips for high-precision ortho generation

• Use cases for ortho images

• Summary

• Frequently asked questions

What is an ortho image?

An ortho image is an image corrected to remove the “distortions” of terrain that appear in aerial photographs, so that it looks as if viewed directly from above. The word “ortho” comes from Greek and means “correct” or “undistorted,” and as the name implies, an ortho image is free of the distortions typical of standard photographs. For example, in the aerial photo layer of the online map services you commonly use, you usually see only rooftops and not building facades. This is because ortho correction (orthographic projection transformation) is applied after shooting to render all objects as if viewed from directly above. Ortho images are also called orthophotos, and because the entire image is represented at a uniform scale like a map, distances and areas can be measured accurately.

Ortho images are normally created not from a single photo but by stitching together multiple aerial images. When shooting a wide area with a drone, the camera’s field of view is limited, so you take dozens of photos and composite them into a single large image using mosaic processing. During this process, positional offsets between photos, lens distortions, and offsets due to terrain undulation are computationally corrected to produce a single ortho mosaic image that looks like a map seen from above. Ortho images have long been used in aerial surveying and geospatial information fields, but with the recent spread of drones it has become easier to obtain high-resolution ortho images.

Benefits of creating ortho images with drones

The growing interest in ortho generation using drones stems from cost reduction and efficiency benefits. Traditionally, creating ortho images over a wide area required aerial photogrammetry with manned aircraft, which involved high costs and long preparation times. Using drones, however, allows detailed aerial photography on site with a small team in a short time. There is no need to rent specialized aircraft or arrange flights, so you can capture images quickly when needed. Also, because drones can fly at low altitudes, they can obtain extremely high-resolution images with ground resolution on the order of a few centimeters (a few in), which is a major advantage.

Improved safety is another benefit. Drones can safely capture the latest conditions from above at dangerous cliffs or disaster sites where people cannot enter. For these reasons, creating ortho images with drones is spreading mainly in civil engineering and construction industries. Local governments are also actively adopting drone aerial surveys for river management and monitoring farmland. With national initiatives promoting construction DX and programs like i-Construction, on-site 3D data capture and ortho image use are becoming standard practice. By combining drones with photogrammetry techniques, map-creation tasks that previously required specialized contractors can increasingly be carried out in-house.

Equipment and software required for ortho generation

Let’s cover what you need to generate ortho images with a drone. First is the drone itself. The higher the camera resolution on the drone, the clearer the ortho image you can obtain. Generally, models equipped with cameras capable of capturing images with several thousand to tens of thousands of pixels are desirable, and many recent commercial drones meet this condition. During shooting, set the camera to point straight down (nadir) so that the ground surface is fully captured. To efficiently cover a wide area, use automatic flight apps to preprogram flight routes and fly in a grid pattern at a consistent altitude and speed.

Next, you need photogrammetry software or a cloud service to process the captured images. Photogrammetry software is the tool that generates ortho images and 3D models (point clouds, DEMs, etc.) from many photos taken by the drone, and it ranges from commercial specialist software to free open-source options. If you use desktop software, a high-performance PC (CPU, GPU, and sufficient memory) makes processing smoother. Recently, however, cloud services have emerged that allow you to upload photos to servers via the internet and automatically receive processed results. Using cloud services lets you process large numbers of photos regardless of your local PC performance, so you can avoid installing dedicated software or investing in hardware.

If you want to improve surveying accuracy, consider using a high-precision GPS (RTK-capable aircraft) or ground survey control points (GCP: Ground Control Point). Ortho images can be generated with a standard drone, but the absolute position accuracy will be affected by the built-in GPS error (on the order of a few meters (a few ft)). By placing targets on the ground with known coordinates in advance and referencing those points during photo processing, you can accurately align the ortho image to real-world coordinates. Likewise, using an RTK-equipped drone enables centimeter-level positioning during flight, improving the positional accuracy of the captured images themselves and reducing post-processing workload. These are not mandatory, but they are worth considering for measurement-focused surveying applications.

Steps to create ortho images

Now let’s look at the general workflow for creating ortho images from drone photos.

• Flight planning and shooting: First, plan the drone flight to cover the target area. For high-quality ortho images, ensuring sufficient overlap between photos is crucial. Generally, 80% or more overlap in the forward direction (flight direction) and 60% or more overlap sideways is recommended. Use an automatic flight program to set altitude and route, and take photos at regular intervals with the camera always pointing at the ground. Pay attention to weather and avoid strong winds and rain; shoot during daylight hours with good visibility. If necessary, place ground markers (GCP targets) to obtain accurate positional information.

• Importing photos and alignment: Load the multiple captured images into photogrammetry software or a cloud service. The software automatically detects feature points in each photo, finds common points (tie points) in overlapping areas, and computes the positional relationships between photos. This estimates each image’s camera position and orientation and reconstructs the three-dimensional structure of the target area (point cloud or 3D model). This process is technically called Structure from Motion (SfM), which reconstructs 3D shape from photos.

• Ortho image generation: Using the reconstructed 3D model, each photo is corrected to a top-down view and stitched together. Specifically, ortho correction projects the photos onto the generated digital elevation model (terrain height information) and seamlessly mosaics adjacent photos. The mosaic processing creates a single continuous ortho image (ortho mosaic). Software typically also automatically adjusts brightness and color so that seams are invisible and the image looks natural. Processing time depends on the number and resolution of photos, but as a guideline, processing a dataset of several hundred images takes a few hours.

• Exporting and using results: Export the completed ortho image in the required format. Commonly, images are exported as GeoTIFF files that include geographic coordinate information and can be loaded into GIS or CAD software. The ortho image can be used on its own as a detailed aerial overview, overlaid with other map data to create current condition maps, or used to prepare survey drawings. For example, an ortho image of a roadworks site can be overlaid on CAD design data to check whether the as-built shape matches the design.

Tips for high-precision ortho generation

To obtain high-quality ortho images, keep the following points in mind.

• Sufficient overlap: As noted above, ensuring plenty of overlap between photos is most important. If overlap is insufficient, the software may fail to align images or cause gaps and distortions. When planning shoots, allow extra overlap—err on the side of more rather than less.

• Nadir shooting: Point the camera as straight down as possible. Oblique photos capture building facades, which makes producing a complete ortho image more difficult. Using only nadir images results in less distortion and higher accuracy. If you must take oblique photos, many workflows allow you to extract and use only the nadir-view images during final processing.

• Lighting and exposure: Pay attention to weather and light conditions. Choose a clear day with direct sunlight when possible. However, strong overhead sunlight at noon can cause specular highlights and blown-out areas or excessively strong contrast. Shooting in the morning or afternoon when the sun is slightly angled can produce clearer object contours and improve 3D reconstruction. Also, instead of relying on automatic exposure, it’s ideal to use manual exposure set uniformly across all photos so brightness does not vary drastically.

• Accuracy improvements: For mapping-grade accuracy, the previously mentioned GCPs and RTK drones are effective. Providing several GCPs with known coordinates as control points in processing allows the entire ortho image to be adjusted to the site’s coordinate system. When using an RTK drone, set up a base station in advance or connect to a network RTK service so positioning during flight is measured accurately; this greatly improves the positional accuracy of each photo. Combining these methods can yield ortho images aligned to centimeter-level accuracy (half-inch accuracy) without further post hoc corrections.

• Compliance with regulations: Compliance with laws such as aviation regulations is essential for drone flights. Flights over urban areas, over people, or large-scale survey flights may require prior permission or approval from the Ministry of Land, Infrastructure, Transport and Tourism. Plan flights and shooting with safety in mind and within the prescribed rules.

Use cases for ortho images

Ortho images generated by drones are used across many fields. Here are representative use cases.

• Surveying for civil engineering and construction sites: Ortho images are used for pre-construction topographic surveys and construction progress control. What used to take survey crews days to produce can now be created quickly with drone aerial photography and ortho generation. Derived ortho images can be used to calculate earthwork volumes or to compare with design drawings for quality control, improving construction management efficiency and sophistication.

• Disaster record and assessment: Ortho images are used for aerial records of disaster-affected areas caused by typhoons or earthquakes. Creating ortho image maps from immediate post-disaster aerial photos helps accurately identify the extent of damage and locations of damaged objects. These maps are easy to share among stakeholders and useful for planning recovery and assessing damages.

• Field management in agriculture: Ortho images created from drone imagery of farmland are used to visualize the overall condition of fields. Farmers can identify uneven crop growth or disease occurrences from a top-down view and reflect that information in fertilizer or pest-control plans as part of precision agriculture (smart farming). Drone aerial photography and ortho generation help farmers and agricultural organizations manage large areas more efficiently.

• Inspection and maintenance of infrastructure: Ortho images assist in inspections of infrastructure such as roads, bridges, and dams. By accumulating ortho images captured during routine surveys and comparing them with past images, signs of deterioration or deformation can be detected early. Managed together with GPS as a base map, they can form a database for multi-year monitoring and maintenance planning.

• Urban planning and public works: Municipalities increasingly use drone ortho images for urban planning and assessing land use. Using high-resolution orthophotos as basemaps enables detection of fine changes in urban areas. There are also cases of publishing ortho image maps of cultural heritage sites or tourist attractions for promotion. Beautiful aerial images captured by drones provide valuable content for public relations materials and educational uses.

Summary

The method of creating ortho images from drone photos is becoming an accessible technique anyone can use. Advances in photogrammetry software and the emergence of cloud services mean that creating orthophotos—once the domain of specialist surveying companies—can now be done with a consumer drone and a PC. High-resolution ortho images combine the accuracy of maps with the intuitive clarity of photographs, making them an indispensable information base across civil engineering, construction, agriculture, and disaster management.

However, when you start doing this yourself you may feel barriers such as the need for a high-performance PC or learning new software. Services are appearing that automatically generate high-precision point clouds and ortho images simply by uploading drone photos to the cloud, similar to simplified surveying with LRTK. Even without specialist knowledge, such user-friendly services and the comprehensive support offered by domestic providers make it easy for beginners to adopt the technology. Use advanced technologies wisely and consider incorporating ortho generation—the new standard—into your work.

Frequently asked questions

Q: What is the difference between an ortho image and a normal photo? A: An ortho image is corrected to appear as if viewed from above, and the entire image is at a uniform scale—this is the main difference from a normal photo. Oblique photos show building facades and have dimensional distortion from perspective, but ortho images are corrected so all objects appear in their top-down shapes, allowing accurate measurement of distances and areas like a map.

Q: Can beginners create ortho images if they have a drone? A: Yes—if you can operate a drone and take basic photos, beginners can create ortho images. Recent automatic flight apps make it easy to set shooting routes, and cloud services and user-friendly software simplify photo analysis. Start with small areas for practice and gradually expand as you gain experience.

Q: Do I need expensive software or special equipment to make ortho images? A: Not necessarily. While specialized photogrammetry software is used to create ortho images, there are free open-source options and cloud services that charge only for use. Cloud services allow processing of large photo sets even without a high-performance PC. Also, commercial consumer drones are usually sufficient, so you can often start with readily available equipment.

Q: How accurate are ortho images made with drones? A: If created following proper procedures, drone ortho images can represent terrain with accuracy of a few centimeters (a few in). However, absolute positional accuracy depends on the GPS and correction methods used. Relying only on a typical commercial drone’s built-in GPS can result in horizontal position errors of several meters (several ft), but using control points (GCPs) or an RTK-capable drone can improve accuracy to within a few centimeters. In any case, the advantage of drone ortho images is the ability to cover wide areas quickly.

Q: In what formats can ortho images be used? A: Ortho images are a type of digital map data and can be used in many ways. They can be loaded into GIS software as basemaps, overlaid in CAD drawings for comparison with existing conditions, and used to measure distances and areas for site surveys, farm management, or post-disaster damage assessment. They can also be printed for reports or overlaid on web map services for public release. Accurate, up-to-date ortho images are widely used as objective materials for decision making.

Q: How many photos do I need to take to create an ortho image? A: The number of photos depends on the area to cover and the flight altitude. For example, with a standard camera on a drone flying at an altitude of 100 m (328.1 ft), a single photo can capture several thousand square meters. Creating an ortho image for several hectares typically requires several hundred photos while maintaining adequate forward and side overlap. Because drone battery capacity is limited, large areas are captured over multiple flights. The more photos you take, the longer processing takes, but cloud services enable efficient analysis of large datasets.